| 1 | If you find inaccuracies in this list, please send mail to |
| 2 | bug-gdb@prep.ai.mit.edu. If you would like to work on any of these, |
| 3 | you should consider sending mail to the same address, to find out |
| 4 | whether anyone else is working on it. |
| 5 | |
| 6 | |
| 7 | Known problems in GDB 5.0 |
| 8 | ========================= |
| 9 | |
| 10 | Below is a list of problems identified during the GDB 5.0 release |
| 11 | cycle. People hope to have these problems fixed in a follow-on |
| 12 | release. |
| 13 | |
| 14 | (The names in paren indicate people that posted the original problem.) |
| 15 | |
| 16 | -- |
| 17 | |
| 18 | GDB doesn't build under IRIX6.4 |
| 19 | |
| 20 | Benjamin Gamsa wrote: |
| 21 | |
| 22 | Has anyone successfully built the latest (from cvs) gdb on IRIX6.4 or |
| 23 | later? The first problem I hit is that proc-api.c includes |
| 24 | sys/user.h, which no longer exists under IRIX6.4. If I comment out |
| 25 | that include, the next problem I hit is that PIOCGETPR and PIOCGETU |
| 26 | are no longer defined in IRIX6.4 (presumably related to the |
| 27 | disappearance of user.h). |
| 28 | |
| 29 | -- |
| 30 | |
| 31 | The BFD directory requires bug-fixed AUTOMAKE et.al. |
| 32 | |
| 33 | AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable. It |
| 34 | contained the full path to texinfo.tex when it should have only |
| 35 | contained the directory. The bug has been fixed in the current |
| 36 | AUTOMAKE sources. Automake snapshots can be found in: |
| 37 | ftp://sourceware.cygnus.com/pub/gdb/snapshots |
| 38 | and ftp://sourceware.cygnus.com/pub/binutils |
| 39 | |
| 40 | -- |
| 41 | |
| 42 | gdb-cvs fails to build on freebsd-elf |
| 43 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00004.html |
| 44 | |
| 45 | Either the FreeBSD group need to contribute their local GDB changes |
| 46 | back to the master sources or someone needs to provides a new |
| 47 | (clean-room) implementation. Since the former involves a fairly |
| 48 | complicated assignment the latter may be easier. [cagney] |
| 49 | |
| 50 | -- |
| 51 | |
| 52 | Generic: lin-thread cannot handle thread exit (Mark Kettenis, Michael |
| 53 | Snyder) http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00525.html |
| 54 | |
| 55 | The thread_db assisted debugging code doesn't handle exiting threads |
| 56 | properly, at least in combination with glibc 2.1.3 (the framework is |
| 57 | there, just not the actual code). There are at least two problems |
| 58 | that prevent this from working. |
| 59 | |
| 60 | As an additional reference point, the pre thread_db code did not work |
| 61 | either. |
| 62 | |
| 63 | -- |
| 64 | |
| 65 | Java (Anthony Green, David Taylor) |
| 66 | |
| 67 | Anthony Green has a number of Java patches that did not make it into |
| 68 | the 5.0 release. |
| 69 | |
| 70 | Patch: java tests |
| 71 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html |
| 72 | |
| 73 | Patch: java booleans |
| 74 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html |
| 75 | |
| 76 | Patch: handle N_MAIN stab |
| 77 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html |
| 78 | |
| 79 | -- |
| 80 | |
| 81 | Pascal (Pierre Muller, David Taylor) |
| 82 | |
| 83 | Pierre Muller has contributed patches for adding Pascal Language |
| 84 | support to GDB. |
| 85 | |
| 86 | 2 pascal language patches inserted in database |
| 87 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html |
| 88 | |
| 89 | Indent -gnu ? |
| 90 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html |
| 91 | |
| 92 | -- |
| 93 | |
| 94 | GNU/Linux/x86 and random thread signals (and Solaris/SPARC but not |
| 95 | Solaris/x86). |
| 96 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00336.html |
| 97 | |
| 98 | Christopher Blizzard writes: |
| 99 | |
| 100 | So, I've done some more digging into this and it looks like Jim |
| 101 | Kingdon has reported this problem in the past: |
| 102 | |
| 103 | http://sourceware.cygnus.com/ml/bug-gdb/1999-10/msg00058.html |
| 104 | |
| 105 | I can reproduce this problem both with and without Tom's patch. Has |
| 106 | anyone seen this before? Maybe have a solution for it hanging around? |
| 107 | :) |
| 108 | |
| 109 | There's a test case for this documented at: |
| 110 | |
| 111 | when debugging threaded applications you get extra SIGTRAPs |
| 112 | http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=9565 |
| 113 | |
| 114 | [There should be a GDB testcase - cagney] |
| 115 | |
| 116 | -- |
| 117 | |
| 118 | Possible regressions with some devel GCCs. |
| 119 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00475.html |
| 120 | |
| 121 | gcc-2.95.2 outputs a line note *before* the prologue (and one for the |
| 122 | closing brace after the epilogue, instead of before it, as it used to |
| 123 | be). By disabling the RTL-style prologue generating mechanism |
| 124 | (undocumented GCC option -mno-schedule-prologue), you get back the |
| 125 | traditional behaviour. |
| 126 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00510.html |
| 127 | |
| 128 | This should now be fixed. |
| 129 | |
| 130 | -- |
| 131 | |
| 132 | RFD: infrun.c: No bpstat_stop_status call after proceed over break? |
| 133 | (Peter Schauer) |
| 134 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html |
| 135 | |
| 136 | GDB misses watchpoint triggers after proceeding over a breakpoint on |
| 137 | x86 targets. |
| 138 | |
| 139 | -- |
| 140 | |
| 141 | x86 linux GDB and SIGALRM (???) |
| 142 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html |
| 143 | |
| 144 | I know there are problems with single stepping through signal |
| 145 | handlers. These problems were present in 4.18. They were just masked |
| 146 | because 4.18 failed to recognize signal handlers. Fixing it is not |
| 147 | easy, and will require changes to handle_inferior_event(), that I |
| 148 | prefer not to make before the 5.0 release. |
| 149 | |
| 150 | Mark |
| 151 | |
| 152 | -- |
| 153 | |
| 154 | Revised UDP support (was: Re: [Fwd: [patch] UDP transport support]) |
| 155 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html |
| 156 | |
| 157 | (Broken) support for GDB's remote protocol across UDP is to be |
| 158 | included in the follow-on release. |
| 159 | |
| 160 | -- |
| 161 | |
| 162 | Can't build IRIX -> arm GDB. |
| 163 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html |
| 164 | |
| 165 | David Whedon writes: |
| 166 | > Now I'm building for an embedded arm target. If there is a way of turning |
| 167 | > remote-rdi off, I couldn't find it. It looks like it gets built by default |
| 168 | > in gdb/configure.tgt(line 58) Anyway, the build dies in |
| 169 | > gdb/rdi-share/unixcomm.c. SERPORT1 et. al. never get defined because we |
| 170 | > aren't one of the architectures supported. |
| 171 | |
| 172 | |
| 173 | -- |
| 174 | |
| 175 | ------------------------------------------------ |
| 176 | |
| 177 | Code cleanups |
| 178 | ============= |
| 179 | |
| 180 | The following code cleanups are planned for the follow-on release to |
| 181 | GDB 5.0. |
| 182 | |
| 183 | -- |
| 184 | |
| 185 | ATTRIBUTE_UNUSED |
| 186 | |
| 187 | The need for this as almost been eliminated. The next version of GCC |
| 188 | (assuming cagney gets the relevant patch committed) will be able to |
| 189 | supress unused parameter warnings. |
| 190 | |
| 191 | -- |
| 192 | |
| 193 | Delete macro TARGET_BYTE_ORDER_SELECTABLE. |
| 194 | |
| 195 | Patches in the database. |
| 196 | |
| 197 | -- |
| 198 | |
| 199 | Updated readline |
| 200 | |
| 201 | Readline 4.? is out. A merge wouldn't hurt. |
| 202 | |
| 203 | -- |
| 204 | |
| 205 | Purge PARAMS |
| 206 | |
| 207 | Eliminate all uses of PARAMS in GDB's source code. |
| 208 | |
| 209 | -- |
| 210 | |
| 211 | Elimination of make_cleanup_func. (Andrew Cagney) |
| 212 | |
| 213 | make_cleanup_func elimination |
| 214 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00791.html |
| 215 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00814.html |
| 216 | |
| 217 | -- |
| 218 | |
| 219 | ChangeLog.mi vs ChangeLog-mi (Andrew Cagney) |
| 220 | Needs further debate. |
| 221 | |
| 222 | Re: [PATCH] Add change-log variables to more MI files |
| 223 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00811.html |
| 224 | |
| 225 | -- |
| 226 | |
| 227 | Re: Various C++ things |
| 228 | |
| 229 | value_headof/value_from_vtable_info are worthless, and should be removed. |
| 230 | The one place in printcmd.c that uses it should use the RTTI functions. |
| 231 | |
| 232 | RTTI for g++ should be using the typeinfo functions rather than the vtables. |
| 233 | The typeinfo functions are always at offset 4 from the beginning of the vtable, |
| 234 | and are always right. The vtables will have weird names like E::VB sometimes. |
| 235 | The typeinfo function will always be "E type_info function", or somesuch. |
| 236 | |
| 237 | value_virtual_fn_field needs to be fixed so there are no failures for virtual |
| 238 | functions for C++ using g++. |
| 239 | |
| 240 | Testsuite cases are the major priority right now for C++ support, since i have |
| 241 | to make a lot of changes that could potentially break each other. |
| 242 | |
| 243 | -- |
| 244 | |
| 245 | GDBARCH cleanup (Andrew Cagney) |
| 246 | |
| 247 | The non-generated parts of gdbarch.{sh,h,c} should be separated out |
| 248 | into arch-utils.[hc]. |
| 249 | |
| 250 | The ``info architecture'' command should be replaced with a fixed |
| 251 | ``set architecture'' (implemented using the command.c enum code). |
| 252 | |
| 253 | Document that gdbarch_init_ftype could easily fail because it didn't |
| 254 | identify an architecture. |
| 255 | |
| 256 | -- |
| 257 | |
| 258 | Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney) |
| 259 | |
| 260 | Add support for packet enable/disable commands with these thread |
| 261 | packets. General cleanup. |
| 262 | |
| 263 | [PATCH] Document the ThreadInfo remote protocol queries |
| 264 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html |
| 265 | |
| 266 | [PATCH] "info threads" queries for remote.c |
| 267 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html |
| 268 | |
| 269 | -- |
| 270 | |
| 271 | Eliminate gdb/tui/Makefile.in. |
| 272 | Cleanup configury support for optional sub-directories. |
| 273 | |
| 274 | Check how GCC handles multiple front ends for an example of how things |
| 275 | could work. A tentative first step is to rationalize things so that |
| 276 | all sub directories are handled in a fashion similar to gdb/mi. |
| 277 | |
| 278 | -- |
| 279 | |
| 280 | [PATCH/5] src/intl/Makefile.in:distclean additions |
| 281 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00363.html |
| 282 | |
| 283 | Do not forget to merge the patch back into the trunk. |
| 284 | |
| 285 | -- |
| 286 | |
| 287 | Re: [RFC] Change configure.in so -W arnings match reality |
| 288 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00350.html |
| 289 | |
| 290 | Some GCC compilers do not like -Wreturn-type. (Going forward there |
| 291 | may be more problems like that). Need to check which of the warning |
| 292 | options are valid. Need to probably disable warnings by default. |
| 293 | |
| 294 | -- |
| 295 | |
| 296 | General Wish List |
| 297 | ================= |
| 298 | |
| 299 | -- |
| 300 | |
| 301 | Check that GDB can handle all BFD architectures (Andrew Cagney) |
| 302 | |
| 303 | There should be a test that checks that BFD/GDB are in sync with |
| 304 | regard to architecture changes. Something like a test that first |
| 305 | queries GDB for all supported architectures and then feeds each back |
| 306 | to GDB.. Anyone interested in learning how to write tests? :-) |
| 307 | |
| 308 | -- |
| 309 | |
| 310 | This list is probably not up to date, and opinions vary about the |
| 311 | importance or even desirability of some of the items. |
| 312 | |
| 313 | Document trace machinery. |
| 314 | |
| 315 | Document overlay machinery. |
| 316 | |
| 317 | Extend .gdbinit mechanism to specify name on command line, allow for |
| 318 | lists of files to load, include function of --tclcommand. |
| 319 | |
| 320 | @c This does not work (yet if ever). FIXME. |
| 321 | @c @item --parse=@var{lang} @dots{} |
| 322 | @c Configure the @value{GDBN} expression parser to parse the listed languages. |
| 323 | @c @samp{all} configures @value{GDBN} for all supported languages. To get a |
| 324 | @c list of all supported languages, omit the argument. Without this |
| 325 | @c option, @value{GDBN} is configured to parse all supported languages. |
| 326 | |
| 327 | Add an "info bfd" command that displays supported object formats, |
| 328 | similarly to objdump -i. |
| 329 | |
| 330 | START_INFERIOR_TRAPS_EXPECTED need never be defined to 2, since that |
| 331 | is its default value. Clean this up. |
| 332 | |
| 333 | It should be possible to use symbols from shared libraries before we know |
| 334 | exactly where the libraries will be loaded. E.g. "b perror" before running |
| 335 | the program. This could maybe be done as an extension of the "breakpoint |
| 336 | re-evaluation" after new symbols are loaded. |
| 337 | |
| 338 | Make single_step() insert and remove breakpoints in one operation. |
| 339 | |
| 340 | Speed up single stepping by avoiding extraneous ptrace calls. |
| 341 | |
| 342 | Speed up single stepping by not inserting and removing breakpoints |
| 343 | each time the inferior starts and stops. |
| 344 | |
| 345 | Breakpoints should not be inserted and deleted all the time. Only the |
| 346 | one(s) there should be removed when we have to step over one. Support |
| 347 | breakpoints that don't have to be removed to step over them. |
| 348 | |
| 349 | Update gdbint.texinfo to include doc on the directory structure and |
| 350 | the various tricks of building gdb. |
| 351 | |
| 352 | Do a tutorial in gdb.texinfo on how to do simple things in gdb. |
| 353 | E.g. how to set a breakpoint that just prints something and continues. |
| 354 | How to break on aborts. Etc. |
| 355 | |
| 356 | Provide "voodoo" debugging of core files. This creates a zombie |
| 357 | process as a child of the debugger, and loads it up with the data, |
| 358 | stack, and regs of the core file. This allows you to call functions |
| 359 | in the executable, to manipulate the data in the core file. |
| 360 | |
| 361 | GDB reopens the source file on every line, as you "next" through it. |
| 362 | |
| 363 | Referencing the vtbl member of a struct doesn't work. It prints OK |
| 364 | if you print the struct, but it gets 0 if you try to deref it. |
| 365 | |
| 366 | Persistent command history: A feature where you could save off a list |
| 367 | of the commands you did, so you can edit it into something that will bring |
| 368 | the target to the same place every time you source it. |
| 369 | This would also be useful for automated fast watchpointing; if you go |
| 370 | past the place where it watchpoints, you just start it over again and |
| 371 | do it more carefully. |
| 372 | |
| 373 | Deal with the SunOS 4.0 and 4.1.1 ptrace bug that loses the registers if |
| 374 | the stack is paged out. |
| 375 | |
| 376 | Finish the C++ exception handling stub routines. Lint points them out |
| 377 | as unused statics functions. |
| 378 | |
| 379 | Perhaps "i source" should take an argument like that of "list". |
| 380 | |
| 381 | See if core-aout.c's fetch_core_registers can be used on more machines. |
| 382 | E.g. MIPS (mips-xdep.c). |
| 383 | |
| 384 | unpack_double() does not handle IEEE float on the target unless the host |
| 385 | is also IEEE. Death on a vax. |
| 386 | |
| 387 | Set up interface between GDB and INFO so that you can hop into interactive |
| 388 | INFO and back out again. When running under Emacs, should use Emacs |
| 389 | info, else fork the info program. Installation of GDB should install |
| 390 | its texinfo files into the info tree automagically, including the readline |
| 391 | texinfo files. |
| 392 | |
| 393 | "help address" ought to find the "help set print address" entry. |
| 394 | |
| 395 | Remove the VTBL internal guts from printouts of C++ structs, unless |
| 396 | vtblprint is set. |
| 397 | |
| 398 | Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if |
| 399 | it matches the source line indicated. |
| 400 | |
| 401 | The prompt at end of screen should accept space as well as CR. |
| 402 | |
| 403 | Check STORE_RETURN_VALUE on all architectures. Check near it in tm-sparc.h |
| 404 | for other bogosities. |
| 405 | |
| 406 | Check for storage leaks in GDB, I'm sure there are a lot! |
| 407 | |
| 408 | vtblprint of a vtbl should demangle the names it's printing. |
| 409 | |
| 410 | Backtrace should point out what the currently selected frame is, in |
| 411 | its display, perhaps showing "@3 foo (bar, ...)" or ">3 foo (bar, |
| 412 | ...)" rather than "#3 foo (bar, ...)". |
| 413 | |
| 414 | "i program" should work for core files, and display more info, like what |
| 415 | actually caused it to die. |
| 416 | |
| 417 | "x/10i" should shorten the long name, if any, on subsequent lines. |
| 418 | |
| 419 | Check through the code for FIXME comments and fix them. dbxread.c, |
| 420 | blockframe.c, and plenty more. (I count 634 as of 940621 - sts) |
| 421 | |
| 422 | "next" over a function that longjumps, never stops until next time you happen |
| 423 | to get to that spot by accident. E.g. "n" over execute_command which has |
| 424 | an error. |
| 425 | |
| 426 | "set zeroprint off", don't bother printing members of structs which |
| 427 | are entirely zero. Useful for those big structs with few useful |
| 428 | members. |
| 429 | |
| 430 | GDB does four ioctl's for every command, probably switching terminal modes |
| 431 | to/from inferior or for readline or something. |
| 432 | |
| 433 | terminal_ours versus terminal_inferior: cache state. Switch should be a noop |
| 434 | if the state is the same, too. |
| 435 | |
| 436 | ptype $i6 = void??! |
| 437 | |
| 438 | Clean up invalid_float handling so gdb doesn't coredump when it tries to |
| 439 | access a NaN. While this might work on SPARC, other machines are not |
| 440 | configured right. |
| 441 | |
| 442 | "b value_at ; commands ; continue ; end" stops EVERY OTHER TIME! |
| 443 | Then once you enter a command, it does the command, runs two more |
| 444 | times, and then stops again! Bizarre... (This behaviour has been |
| 445 | modified, but it is not yet 100% predictable when e.g. the commands |
| 446 | call functions in the child, and while there, the child is interrupted |
| 447 | with a signal, or hits a breakpoint.) |
| 448 | |
| 449 | help completion, help history should work. |
| 450 | |
| 451 | Check that we can handle stack trace through varargs AND alloca in same |
| 452 | function, on 29K. |
| 453 | |
| 454 | wait_for_inferior loops forever if wait() gives it an error. |
| 455 | |
| 456 | "i frame" shows wrong "arglist at" location, doesn't show where the args |
| 457 | should be found, only their actual values. |
| 458 | |
| 459 | There should be a way for "set" commands to validate the new setting |
| 460 | before it takes effect. |
| 461 | |
| 462 | A mess of floating point opcodes are missing from sparc-opcode.h. |
| 463 | Also, a little program should test the table for bits that are |
| 464 | overspecified or underspecified. E.g. if the must-be-ones bits |
| 465 | and the must-be-zeroes bits leave some fields unexamined, and the format |
| 466 | string leaves them unprinted, then point this out. If multiple |
| 467 | non-alias patterns match, point this out too. Finally, there should |
| 468 | be a sparc-optest.s file that tries each pattern out. This file |
| 469 | should end up coming back the same (modulo transformation comments) |
| 470 | if fed to "gas" then the .o is fed to gdb for disassembly. |
| 471 | |
| 472 | Eliminate all the core_file_command's in all the xdep files. |
| 473 | Eliminate separate declarations of registers[] everywhere. |
| 474 | |
| 475 | "ena d" is ambiguous, why? "ena delete" seems to think it is a command! |
| 476 | |
| 477 | Perhaps move the tdep, xdep, and nat files, into the config |
| 478 | subdirectories. If not, at least straighten out their names so that |
| 479 | they all start with the machine name. |
| 480 | |
| 481 | inferior_status should include stop_print_frame. It won't need to be |
| 482 | reset in wait_for_inferior after bpstat_stop_status call, then. |
| 483 | |
| 484 | i line VAR produces "Line number not known for symbol ``var''.". I |
| 485 | thought we were stashing that info now! |
| 486 | |
| 487 | We should be able to write to random files at hex offsets like adb. |
| 488 | |
| 489 | Make "target xxx" command interruptible. |
| 490 | |
| 491 | Handle add_file with separate text, data, and bss addresses. Maybe |
| 492 | handle separate addresses for each segment in the object file? |
| 493 | |
| 494 | Handle free_named_symtab to cope with multiply-loaded object files |
| 495 | in a dynamic linking environment. Should remember the last copy loaded, |
| 496 | but not get too snowed if it finds references to the older copy. |
| 497 | |
| 498 | Generalize and Standardize the RPC interface to a target program, |
| 499 | improve it beyond the "ptrace" interface, and see if it can become a |
| 500 | standard for remote debugging. (This is talking about the vxworks |
| 501 | interface. Seems unlikely to me that there will be "a standard" for |
| 502 | remote debugging anytime soon --kingdon, 8 Nov 1994). |
| 503 | |
| 504 | Remove all references to: |
| 505 | text_offset |
| 506 | data_offset |
| 507 | text_data_start |
| 508 | text_end |
| 509 | exec_data_offset |
| 510 | ... |
| 511 | now that we have BFD. All remaining are in machine dependent files. |
| 512 | |
| 513 | When quitting with a running program, if a core file was previously |
| 514 | examined, you get "Couldn't read float regs from core file"...if |
| 515 | indeed it can't. generic_mourn_inferior... |
| 516 | |
| 517 | Have remote targets give a warning on a signal argument to |
| 518 | target_resume. Or better yet, extend the protocols so that it works |
| 519 | like it does on the Unix-like systems. |
| 520 | |
| 521 | Sort help and info output. |
| 522 | |
| 523 | Re-organize help categories into things that tend to fit on a screen |
| 524 | and hang together. |
| 525 | |
| 526 | renote-nindy.c handles interrupts poorly; it error()s out of badly |
| 527 | chosen places, e.g. leaving current_frame zero, which causes core dumps |
| 528 | on the next command. |
| 529 | |
| 530 | Add in commands like ADB's for searching for patterns, etc. We should |
| 531 | be able to examine and patch raw unsymboled binaries as well in gdb as |
| 532 | we can in adb. (E.g. increase the timeout in /bin/login without source). |
| 533 | |
| 534 | Those xdep files that call register_addr without defining it are |
| 535 | probably simply broken. When reconfiguring this part of gdb, I could |
| 536 | only make guesses about how to redo some of those files, and I |
| 537 | probably guessed wrong, or left them "for later" when I have a |
| 538 | machine that can attempt to build them. |
| 539 | |
| 540 | When doing "step" or "next", if a few lines of source are skipped between |
| 541 | the previous line and the current one, print those lines, not just the |
| 542 | last line of a multiline statement. |
| 543 | |
| 544 | When searching for C++ superclasses in value_cast in valops.c, we must |
| 545 | not search the "fields", only the "superclasses". There might be a |
| 546 | struct with a field name that matches the superclass name. This can |
| 547 | happen when the struct was defined before the superclass (before the |
| 548 | name became a typedef). |
| 549 | |
| 550 | Handling of "&" address-of operator needs some serious overhaul |
| 551 | for ANSI C and consistency on arrays and functions. |
| 552 | For "float point[15];": |
| 553 | ptype &point[4] ==> Attempt to take address of non-lvalue. |
| 554 | For "char *malloc();": |
| 555 | ptype malloc ==> "char *()"; should be same as |
| 556 | ptype &malloc ==> "char *(*)()" |
| 557 | call printf ("%x\n", malloc) ==> weird value, should be same as |
| 558 | call printf ("%x\n", &malloc) ==> correct value |
| 559 | |
| 560 | Fix dbxread.c symbol reading in the presence of interrupts. It |
| 561 | currently leaves a cleanup to blow away the entire symbol table when a |
| 562 | QUIT occurs. (What's wrong with that? -kingdon, 28 Oct 1993). |
| 563 | |
| 564 | Mipsread.c reads include files depth-first, because the dependencies |
| 565 | in the psymtabs are way too inclusive (it seems to me). Figure out what |
| 566 | really depends on what, to avoid recursing 20 or 30 times while reading |
| 567 | real symtabs. |
| 568 | |
| 569 | value_add() should be subtracting the lower bound of arrays, if known, |
| 570 | and possibly checking against the upper bound for error reporting. |
| 571 | |
| 572 | mipsread.c symbol table allocation and deallocation should be checked. |
| 573 | My suspicion is that it's full of memory leaks. |
| 574 | |
| 575 | SunOS should have a target_lookup_symbol() for common'd things allocated |
| 576 | by the shared library linker ld.so. |
| 577 | |
| 578 | When listing source lines, check for a preceding \n, to verify that |
| 579 | the file hasn't changed out from under us. |
| 580 | |
| 581 | When listing source lines, eat leading whitespace corresponding to the |
| 582 | line-number prefix we print. This avoids long lines wrapping. |
| 583 | |
| 584 | mipsread.c needs to check for old symtabs and psymtabs for the same |
| 585 | files, the way it happens for dbxread.c and coffread.c, for VxWorks |
| 586 | incremental symbol table reloading. |
| 587 | |
| 588 | Get all the remote systems (where the protocol allows it) to be able to |
| 589 | stop the remote system when the GDB user types ^C (like remote.c |
| 590 | does). For ebmon, use ^Ak. |
| 591 | |
| 592 | Possible feature: A version of the "disassemble" command which shows |
| 593 | both source and assembly code ("set symbol-filename on" is a partial |
| 594 | solution). |
| 595 | |
| 596 | investigate "x/s 0" (right now stops early) (I think maybe GDB is |
| 597 | using a 0 address for bad purposes internally). |
| 598 | |
| 599 | Make "info path" and path_command work again (but independent of the |
| 600 | environment either of gdb or that we'll pass to the inferior). |
| 601 | |
| 602 | Make GDB understand the GCC feature for putting octal constants in |
| 603 | enums. Make it so overflow on an enum constant does not error_type |
| 604 | the whole type. Allow arbitrarily large enums with type attributes. |
| 605 | Put all this stuff in the testsuite. |
| 606 | |
| 607 | Make TYPE_CODE_ERROR with a non-zero TYPE_LENGTH more useful (print |
| 608 | the value in hex; process type attributes). Add this to the |
| 609 | testsuite. This way future compilers can add new types and old |
| 610 | versions of GDB can do something halfway reasonable. |
| 611 | |
| 612 | Clean up formatting of "info registers" on MIPS and 88k. See if it |
| 613 | is possible to do this generically across all target architectures. |
| 614 | |
| 615 | GDB gets bfd/corefile.c and gdb/corefile.c confused (this should be easy to |
| 616 | repeat even with something more recent than GDB 4.9). |
| 617 | |
| 618 | Check that unmatched RBRAC doesn't abort(). |
| 619 | |
| 620 | Fix mdebugread.c:parse_type to do fundamental types right (see |
| 621 | rs6000_builtin_type in stabsread.c for what "right" is--the point is |
| 622 | that the debug format fixes the sizes of these things and it shouldn't |
| 623 | depend on stuff like TARGET_PTR_BIT and so on. For mdebug, there seem |
| 624 | to be separate bt* codes for 64 bit and 32 bit things, and GDB should |
| 625 | be aware of that). Also use a switch statement for clarity and speed. |
| 626 | |
| 627 | Investigate adding symbols in target_load--some targets do, some |
| 628 | don't. |
| 629 | |
| 630 | Put dirname in psymtabs and change lookup*symtab to use dirname (so |
| 631 | /foo/bar.c works whether compiled by cc /foo/bar.c, or cd /foo; cc |
| 632 | bar.c). |
| 633 | |
| 634 | Merge xcoffread.c and coffread.c. Use breakpoint_re_set instead of |
| 635 | fixup_breakpoints. |
| 636 | |
| 637 | Fix byte order and int size sins in tm-a29k.h |
| 638 | (EXTRACT_RETURN_VALUE). Perhaps should reproduce bug and verify fix |
| 639 | (or perhaps should just fix it...). |
| 640 | |
| 641 | Make a watchpoint on a constant expression an error (or warning |
| 642 | perhaps) |
| 643 | |
| 644 | Make a watchpoint which contains a function call an error (it is |
| 645 | broken now, making it work is probably not worth the effort). |
| 646 | |
| 647 | Re-do calls to signal() in remote.c, and inflow.c (set_sigint_trap and |
| 648 | so on) to be independent of the debugging target, using target_stop to |
| 649 | stop the inferior. Probably the part which is now handled by |
| 650 | interrupt_query in remote.c can be done without any new features in |
| 651 | the debugging target. |
| 652 | |
| 653 | New test case based on weird.exp but in which type numbers are not |
| 654 | renumbered (thus multiply defining a type). This currently causes an |
| 655 | infinite loop on "p v_comb". |
| 656 | |
| 657 | Nuke baseclass_addr. |
| 658 | |
| 659 | Nuke USG define. |
| 660 | |
| 661 | "source file more recent" loses on re-read |
| 662 | |
| 663 | Fix 386 floating point so that floating point registers are real |
| 664 | registers (but code can deal at run-time if they are missing, like |
| 665 | mips and 68k). This would clean up "info float" and related stuff. |
| 666 | |
| 667 | Look at Solaris bug in interrupt.exp. Can get out of syscall with |
| 668 | PRSABORT (syscall will return EINTR) but merely doing that leads to a |
| 669 | "can't read memory" error. |
| 670 | |
| 671 | gcc -g -c enummask.c then gdb enummask.o, then "p v". GDB complains |
| 672 | about not being able to access memory location 0. |
| 673 | |
| 674 | -------------------- enummask.c |
| 675 | enum mask |
| 676 | { |
| 677 | ANIMAL = 0, |
| 678 | VEGETABLE = 1, |
| 679 | MINERAL = 2, |
| 680 | BASIC_CATEGORY = 3, |
| 681 | |
| 682 | WHITE = 0, |
| 683 | BLUE = 4, |
| 684 | GREEN = 8, |
| 685 | BLACK = 0xc, |
| 686 | COLOR = 0xc, |
| 687 | |
| 688 | ALIVE = 0x10, |
| 689 | |
| 690 | LARGE = 0x20 |
| 691 | } v; |
| 692 | |
| 693 | If try to modify value in file with "set write off" should give |
| 694 | appropriate error not "cannot access memory at address 0x65e0". |
| 695 | |
| 696 | Why do we allow a target to omit standard register names (NO_STD_REGS |
| 697 | in tm-z8k.h)? I thought the standard register names were supposed to |
| 698 | be just that, standard. |
| 699 | |
| 700 | Allow core file without exec file on RS/6000. |
| 701 | |
| 702 | Make sure "shell" with no arguments works right on DOS. |
| 703 | |
| 704 | Make gdb.ini (as well as .gdbinit) be checked on all platforms, so |
| 705 | the same directory can be NFS-mounted on unix or DOS, and work the |
| 706 | same way. |
| 707 | |
| 708 | cd ~/tmp/<M-?> causes infinite loop (where ~/tmp is a directory). |
| 709 | |
| 710 | Get SECT_OFF_TEXT stuff out of objfile_relocate (might be needed to |
| 711 | get RS/6000 to work right, might not be immediately relevant). |
| 712 | |
| 713 | Clean up add_toc_to_loadinfo |
| 714 | |
| 715 | Think about attached processes and sharing terminal. |
| 716 | |
| 717 | John sez in reference to ignoring errors from tcsegpgrp if attach_flag: |
| 718 | set_tty_state should not have any trouble with attached processes. |
| 719 | Instead, the tty handling should leave the pgrp of the tty alone when |
| 720 | attaching to processes (perhaps pass terminal_init_inferior a flag |
| 721 | saying whether we're attaching). |
| 722 | |
| 723 | PAGE_SIZE redefined warnings on AIX. Probably should be using |
| 724 | BFD_PAGE_SIZE throughout BFD. |
| 725 | |
| 726 | Rewrite proceed, wait_for_inferior, and normal_stop to clean them up. |
| 727 | Suggestions: |
| 728 | |
| 729 | 1) Make each test in wait_for_inferior a seperate subroutine |
| 730 | call. |
| 731 | 2) Combine wait_for_inferior and normal_stop to clean up |
| 732 | communication via global variables. |
| 733 | 3) See if you can find some way to clean up the global |
| 734 | variables that are used; possibly group them by data flow |
| 735 | and information content? |
| 736 | |
| 737 | Work out some kind of way to allow running the inferior to be done as |
| 738 | a sub-execution of, eg. breakpoint command lists. Currently running |
| 739 | the inferior interupts any command list execution. This would require |
| 740 | some rewriting of wait_for_inferior & friends, and hence should |
| 741 | probably be done in concert with the above. |
| 742 | |
| 743 | Add function arguments to gdb user defined functions. |
| 744 | |
| 745 | Add convenience variables that refer to exec file, symbol file, |
| 746 | selected frame source file, selected frame function, selected frame |
| 747 | line number, etc. |
| 748 | |
| 749 | Add a "suspend" subcommand of the "continue" command to suspend gdb |
| 750 | while continuing execution of the subprocess. Useful when you are |
| 751 | debugging servers and you want to dodge out and initiate a connection |
| 752 | to a server running under gdb. |
| 753 | |
| 754 | Add stab information to allow reasonable debugging of inline functions |
| 755 | (possibly they should show up on a stack backtrace? With a note |
| 756 | indicating that they weren't "real"?). |
| 757 | |
| 758 | Modify the naked "until" command to step until past the current source |
| 759 | line, rather than past the current pc value. This is tricky simply |
| 760 | because the low level routines have no way of specifying a multi-line |
| 761 | step range, and there is no way of saying "don't print stuff when we |
| 762 | stop" from above (otherwise could just call step many times). |
| 763 | |
| 764 | Modify the handling of symbols grouped through BINCL/EINCL stabs to |
| 765 | allocate a partial symtab for each BINCL/EINCL grouping. This will |
| 766 | seriously decrease the size of inter-psymtab dependencies and hence |
| 767 | lessen the amount that needs to be read in when a new source file is |
| 768 | accessed. |
| 769 | |
| 770 | Do an "x/i $pc" after each stepi or nexti. |
| 771 | |
| 772 | Modify all of the disassemblers to use printf_filtered to get correct |
| 773 | more filtering. |
| 774 | |
| 775 | Modify gdb to work correctly with Pascal. |
| 776 | |
| 777 | Add a command for searching memory, a la adb. It specifies size, |
| 778 | mask, value, start address. ADB searches until it finds it or hits |
| 779 | an error (or is interrupted). |
| 780 | |
| 781 | Remove the range and type checking code and documentation, if not |
| 782 | going to implement. |
| 783 | |
| 784 | # Local Variables: |
| 785 | # mode: text |
| 786 | # End: |